When dropping collision makes sense
Here is the rule first, then three cars to try it on. If it holds up for all three, you can try it on yours.
Collision is worth keeping while the most it could ever pay you is more than five years of what it costs.
ceiling = car's value − deductible
years of premium = ceiling ÷ annual collision premium
under 5 → give dropping it a serious look
The logic is that collision coverage has a hard ceiling. If the car is wrecked, the insurer pays roughly what the car was worth before the crash, less your deductible, and not a dollar more. As the car ages, that ceiling falls every year. The premium usually falls too, but more slowly, because repair costs don't shrink with the car's value. At some point you are paying a large share of the ceiling each year for the right to claim it.
Five years is our choice of threshold, not an industry standard. It reflects a plain judgment: if you would need five or more claim-free years of premiums to fund one worst case yourself, the insurance is still doing a job that is hard to do alone. Below that, a disciplined savings cushion starts to compete. Choose a different number if your own circumstances argue for it; the method still works.
Two conditions sit ahead of the rule and override it. First, if the car is financed or leased, the contract will usually require collision coverage (check yours), and the rule doesn't apply until the lender is paid. Second, if losing the car with no payout would leave you unable to get to work, keep the coverage regardless of what the arithmetic says. The rule is for people who could absorb the loss.
Now the trials. The car values, premiums and the $1,000 deductible are all illustrative, chosen to show the rule at different points in a car's life. Your insurer can tell you exactly what your collision coverage costs as a separate line; it is itemized on the declarations page.
Car one a few years old
Ceiling: $16,000 − $1,000 = $15,000. Years of premium: $15,000 ÷ $660 = 22.7.
Not close. It would take more than two decades of premiums to add up to the one payout the coverage could make. Even a household with a healthy emergency fund would be taking on a large risk to save $660 a year. The coverage is doing exactly what insurance is for: turning a large, unlikely loss into a small, certain cost.
Car two getting older
Ceiling: $6,500 − $1,000 = $5,500. Years of premium: $5,500 ÷ $520 = 10.6.
The car has lost more than half its value since trial one, while the premium dropped by only about a fifth, from $660 to $520. That mismatch is the pattern the rule is designed to catch. Ten years is still well above the line, so collision stays. But this is the stage where the check becomes an annual habit, because the number is now falling fast.
Car three old and paid off
Ceiling: $2,800 − $1,000 = $1,800. Years of premium: $1,800 ÷ $450 = 4.0.
Now the rule fires. Four years of premium, $1,800, equals the most the coverage could ever pay out on this car. Put another way, if you set aside the $450 each year instead, after four claim-free years you would be holding the whole payout yourself, and every year after that is money the insurer would never have returned. If you have a crash in year two instead, you would be $900 into your own fund and $900 short of what the coverage would have paid, a real but bounded loss on a car worth $2,800.
Lowering the deductible doesn't rescue it. At a $500 deductible the ceiling rises to $2,300, but the premium rises as well; at an illustrative $540 a year, $2,300 ÷ $540 is still only 4.3 years.
When to run it
Once a year, when the renewal arrives, is enough. The two inputs that matter both change on roughly that schedule: the car's value slides a little further, and the renewal shows the new collision premium on its own line. Write the result on the renewal notice itself. After a few years the numbers form a short series, and you can watch the figure approach five instead of discovering one day that it passed it long ago.
What the trials leave out
Comprehensive is a separate decision. It covers theft, weather, glass and animal strikes, is often cheaper than collision, and can be kept after collision goes. Run the same rule on its own premium.
Your own car's value, honestly. Use what the car would actually sell for in its present condition, not what you paid or what you hope. What a total loss would pay out depends on how the insurer arrives at actual cash value, which what your insurer pays when your car is totaled walks through.
The money has to go somewhere. Dropping collision only works if the saved premium lands in a fund you won't touch. That fund is the thing you are replacing the insurance with, and sizing a car repair fund covers how large to make it. Without it, you haven't self-insured; you've just uninsured.
Liability stays. None of this touches the coverage that pays other people. An old car can cause a very expensive crash, and dropping collision is not a reason to lower liability limits. The deductible arithmetic in choosing a deductible is the other lever on an aging car's bill, and it is often worth pulling first.
Keep reading
Further reading
- Collision insurance Investopedia
- Auto insurance basics Triple-I
- Auto insurance resources Triple-I